periphrastic verb phrases in spanish
TRANSCRIPT
Periphrastic Verb Phrases in Spanish:
A Glimpse of Some Semantic and Syntactic Properties
Cara Moye
Swarthmore College
April 16, 1999
Title page
Table of Contents
1. Introduction
II. Background Information
Table of Contents
A. Reasons for Using and Examining Periphrastics
B. When and Where do They Occur
C. Distinguishing Periphrastic Phrases
D. Translating Periphrasties
III. V IP Structures
A. Lexical Entries and Sentence Structures
B. ClWe Cllmbing in VIP's
C. Wh- Movement in VIP's
D. Passive Movement in V lP's
IV. V CP Structures
A. Lexical Entries and Sentence Structures
B. CHtic Climbing in V CP's
C. Wb- Movement in V CP's
D. Passive Movement in V CP's
V. Conclusion
Bibliography
1
2
3
4
4
5
10
16
24
26
30
35
37
43
44
48
52
S6
60
61
2
I. Introduction l
In Spanish, there exists a verbal construction such that there are at least two verbs,
and the first verb modifies the second verb. These structures vary slightly in their
construction, from the V IP structure of anda vestido (goes around dressed? or anda
mefiindose (goes around messing) to the V CP construction of llega a verla (to actually
see her. Nonetheless, these two types of structures act much alike in their semantic
makeup. First, we will explore them in non-technical terms, seeing how linguists
explored them in the past, and looking at their constructions using descriptive methods as
opposed to generative grammar methods. In this section, we will explore what they are,
what kinds of verbs are used, and where and when they occur in the language. We will
also note some problems that can occur in translation of these verb phrases, giving a
speciflc example. After this non-technical background, the following sections will focus
on the syntax of these structures and how Govenunent and Binding Theory accounts for
them. In these sections, we will look at the clausal V IP structure that is comprised of the
Verb + Past Participle and the Verb + NDO constructions. Then we will look at more
complicated data, which includes the V CP construction and problems that occur in GB
Theory with these structures.
I Acknowledgements: I would sincerely like to thank Ted Fernald, Donna Jo Napoli, Kari Swingle, Charles O'Brien, Mar Campelos, Jesus Fernandez Alvarez, Justin Aglialoro, Keith Early, Ellen Johnson, Carmen Bender, Cathy Bender, Arantxa Gil Irizar, Jill Ratzan, Brenda Baxter and Claudia Munoz for their suggestions, endless proof-reading, support, and help. Any errors which doubtless remain are, of course, mme. 2 All translations are my own, unless otherwise specified.
3
II. Background Information
A. Reasons for Using and Examining Periphrastics
Spanish periphrastic phrases have become a very important topic in Spanish
linguistics, due to their extensive use throughout the written and spoken language. The
studies of this verbal phenomenon have been ongoing in Spanish linguistics since the
early 1900's. In earlier texts, such as Samuel Gili y Gaya's descriptive syntactic text
(1943), we find the basis for many arguments today. The tendency in Spain is to steer
toward semantic descriptions and translation techniques, though the syntactic properties
of theses PV's are also interesting. They properties differentiate them from regular verb
phrases, as we will see.
It has been said that Spanish speakers use the PV's so often because adverbs are
not so commonly used in the language. This is because Spanish js such a
rythm-conscious language, and adverbs can be extremely long. They are usually fanned
by adding -mente (two syllables) to the end of an adjective (which itself can sometimes
be 3 or 4 syllables.) For instance, an adjective like evidente (which means evident and
has 4 syllables) with the addition of -mente, becomes evidentemente (6 syllables.) The
length of adverbs can sound very strange, or as I was told by a native Spanish speaker
from Spain, it sounds "cargado," or stuffy. In addition, this can really break up rhythm,
so speakers avoid using it whenever possible.
One manner of getting this meaning without using an adverb is by periphrastic
verb phrases (pV's.) Perhaps because long adverbs are not avoided in languages such as
4
English (and French) the tendency to have special verb phrases like these is much tess
evident. Fewer PV's exists in these languages. English and French, for instance, both
share a common periphrastic phrase (in meaning) with Spanish: to be going + infinitive,
aller + infinitive, ir + a + infinitive. The syntactic structures of the 3 languages are
different, but in all of them the meaning of some intention in the future is added to the
infinitive verb. However, Spanish has over 60 of these structures, many more than
English and French together.
B. Where and When Do They Occur?
The use of these verb phrases, as mentioned before, is really quite common in
Spain. It ranges from very colloquial (andar -I- NDO or PP) to a very eloquent, poetic
speech (haber de + infinitive.) Social contexts do influence the choice of using a
periphrastic phrase or not. Some periphrastic verbs modify the second verb in a way
which has nothing or very little to do with their non-periphrastic meaning (while others
take on a more figurative sense.) For instance, in a phrase like 1, given below, andar
really does not carry its meaning to walk at all, but rather gives the impression that the
speaker does not approve of the action, and that the doer of the action is doing something
wrong. I have balded the periphrastic phrase to call attention to it:
1. Este chiC<) anda meti{ndosc con Ja gente.
This boy walks messing (reflexive) with people.
TIlis boy goes around messing with people.
5
In this phrase, the boy does not walk. He isn't necessarily going anywhere. The
speaker of the sentence has a pejorative concept of what the boy is doing: messing with
people. Andar is modifying the meaning of meterse. You could also say this phrase as in
sentence 2, without the verb andar, but it loses the displeasure on the part of the speaker:
2. Este chico se mete con la gente.
This boy (reflexive) messes with people.
In fact, sentence 2, while not really eloquent due to the more colloquial usage of
meterse (which can also mean to put or to place, and can itself be used periphrastically)
would probably be more acceptable in a rigid social environment than sentence 1. The
use of andar in periphrastic phrases is very colloquial. The change in meaning is evident
from the examples above. In sentence 1, we can certainly note that the meaning of andar
is not to walk, although in nonnal circumstances, and in all social environments, andar
means to walk. Interestingly enough, however, this verb is commonly used in many
colloquial expressions. For instance, a young Colombian friend of mine once used the
expreSSiOn:
3. Esta nina y yo, andamos muy bien.
Literally: This girl and I, we walk very well.
This girl and 1 get along really well.
Andar in the you infonna1 command fonn is often used as an exclamation that
something another speaker said or did is ridiculous, questionable, or not beJieved at all.
A professor in Spain once said to me, incredulously because I was carrying a tiny
bilingual dictionary (which she deemed inadequate):
6
4. iAy Dies mio! Anda! LY r6 te lIevas ese diccionario?
Oh my lord! Get out! And you acrually carry around this dictionary?
So andar does have a tendency to be used in colloquial situations, and it does
have a tendency to stray from its more formal meaning of to walk. There are many other
verbs used in periphrastic phrases, of course. And the degree to which they stray from
their non-periphrastic meaning varies. For instance, in a non-formal class setting, one can
find periphrastic phrases such as no acabar de + infinitive. I have included no in this
phrase because it adds frustration that does not occur when no is not present. In
non-periphrastic phrases the verb acabar means to finish. In non-negative periphrastic
phrases, acabar de modifies the second verb, adding that the action of the second verb
has just been completed. The negative of acabar de adds frustration that the action has
not been completed entirely. A Spanish professor in class once addressed a student's
shyness to repeat a question:
5. No! jRepitelo! Es que no acabo de entender la primera parte.
No! Repeat it! It's that I just didn't understand the first part!
Here again the periphrastic phrase is in bold. The verb acabar in a
non-periphrastic phrase can be used in any social context. Periphrastically it is possible
that one would not find it in a very formal social setting. A non-periphrastic use of
acabar is given in sentence 6:
6. Se acab6 la Iluvia
The rain has just stopped.
7
Acabar is a synonym for the verb terminar, which also means 10 finish. However)
a phrase like 7 can be generated without redundancy:
7. Acabo de terminar.
I have just finished.
This is because acabar no longer means to finish and is no longer interchangeable
with lerminar. Terminar cannot be used in a periphrastic phrase, and the two verbs here
cannot be reversed. Thus, you would never have a sentence like 8 generated:
8 * Termino de acabar
I finish of to finish .
Another common and rather colloquial periphrastic phrase is meterse a -I-
infinitive. As the second verb of a periphrastic verb phrase used before in example 1, we
can see that this verb is very flexible in its colloquial meanings. The meaning it adds to
its second verb is one of initiation of an action, but it also adds a rather pejorative aspect
to the action. It gives an impression that the speaker judges that the doer of the action
does not really know what he or she is doing, or that the doer is not capable of doing the
action. While it can be a bit difficult to apply this translation in English, an example of
this verb in a periphrastic phrase is given below in sentence 9. Because it is the speaker
himself who is determining his own action in a pejorative manner, it is especially difficult
to get the full meaning of the periphrastic phrase when translating to English. A man in a
train said to an acquaintance in that train:
9 Puedes jrte cuando quieras. No se: .. Es que me meto a hablar y ya veraS ...
You can go whenever you like. I don't know. It's that T get wrapped up in
talking and .. . you' 11 see.
8
Meterse, in its formal usage, as mentioned earlier means to put or to place. Not
only can it be used in fonnal contexts, but it is a more highly descriptive word than its
synonymponerse, which also means to put or to place. Ponerse, additionally, can be
used in periphrastic verb phrases. Another synonym, which is deemed even more
eloquent and more fonnal than either of these two is the verb coiocar, and it cannot be
used as a periphrastic verb. This will become more important later when we look at the
semantic tendencies of peri phrastic verbs.
All of the former examples of periphrastic phrases have been in an at least
somewhat colloquial envirorunent. However, as mentioned before, some of them do
occur frequently in very formal speech too. In a political and economic book, As{
CambiarcfEspana, the writer addresses his audience in a very fonnal setting, and quotes
several politicians addressing their pubJic. One politician (Pierre \flimlun) was quoted as
saymg:
]0 .. .1a unificacion de Europa es una idea justa y por ello acabad triunfando .
... the unification of Europe is ajust idea and because of that, it will end up triumphant
Again we already know that the verb acabar means fo finish, and that it is used to
mean fo have just when used with de and an infinitive. In the phrase above it is used with
the -NDO (the progressive) form of the verb triumph, and is not colloquial at all. Also, it
has only undergone a slight alteration of meaning, since it still carries a meaning of to
finish or end in both periphrastic phrases.
9
Thus, we can draw from the examples above that the PV's do occur in all levels of
the language, but that certain periphrastic phrases are usable in colloquial contexts only.
We have also seen a glimpse of how verbs that are used often in colloquial contexts are
often used in expressions, and are flexible verbs in that they can carry a variety of
meanings depending on the context, can be used periphrastically.
C. Distinguishing Periphrastic Phrases
In this section, we will mostly look at the different arguments among linguists
today about what does and does not make up a periphrastic verb phrase. In addition, we
will see some interesting occurrences and uses of periphrastic phrases. According to
many linguistics, one of the requisites that a verb phrase must fulfill in order to be called
periphrastic is the loss of the original meaning of the conjugated verb. That is to say, the
conjugated verb modifies the second verb, and not with the meaning it carries when used
alone. One standard way of saying this was to say that the first verb acts as an auxiliary
verb. According to Samuel GilL y Gaya (1943, p. 95), "para distinguir si un verbo esta'
empleado como auxiliar basta fijarse en si ha perdido su significado propio." (In order to
determine if a verb is employed as an auxiliary, all that is needed is to look at whether or
not it has lost its own meaning.) This is still one method used to identify peliphrastic
phrases.
The real problem that occurs with this type of explanation is that there are verb
phrases in which the first verb does not really lose its original meaning, but rather
modifies the second verb with the meaning that it carries when used alone. If the above
10
explanation of semantic tendencies is used, than these verb plrrases cannot be considered
periphrastic even though they act exactly the same syntactically. This is probably not the
most accurate manner of analyzing these phrases, since ideally there would not be a
whole different type of phrase that follows the exact same syntactic patterns, but differ in
whether the first verb retains or loses meaning. There are other linguists, such as Rafael
Gomez Fente, Jesus FermlD.dez Alvarez, and Lope G. Feijoo (1987), that claim there are
periplrrastic verb phrases and 'medio-perifrasis' (or haIf periphrastic phrases.) Those
whose first verbs do not lose their meaning, but follow all ofthe syntactic patterns of a
PV, are considered 'medio-perifrasis.' This line ofthinking does not really offer much
more than that of Gili y Gaya, though it does offer some line between them semantically
without excluding the syntactically. However, they do not function as haIf-periphIastic
phrases syntactically, and calling them such may suggest that. There really is not a need
for it; mainly for tl1is reason~ there are others, such as Hella Olbertz (1998), that argue
that all periphrastic phrases that meet the special syntactic tests are indeed PV's, and that
the loss of original meaning is just special to some phrases.
If we accept this line of reasoning, looking at the semantic properties of the
periphrastic verb phrases we can find a whole range of verbs that vary from a complete
loss of their non-periphrastic meaning, to verbs that have the same sense both
periphrastically, and non-periphrastically. On the following page is a chart of verbs that
can be used periphrastically with both their non-periphrastic and their periphrastic
meanings (the given periphrastic meanings below are my translations of the meanings
given in Fente-Go'mez, 1987):
11
Table 1 - Periphrastic Verbs with Periphrastic and Non-periphrastic Meanings
echarse a+ infinitive (0 lie down, to throw oneself to burst out
~char a + infinilive (0 throw, to throw out, to post, to burst out/take off
to sack, to give off
volver a + infinitive to retum repetition, again ,
liarse a + infinitive to get muddled up, involved to get involved, forget everything else
hincharse de + infinitive to inflate, exaggerate to overdo it, get one's fill of
dejar de + infinitive to leave to quit, stop
lacabar de+ infinitive to end, to finish to have just
acabar por I- infinitive to end, to finish to end up
seguir + NOO to continue 10 continue, keep on
pasar a + infinitive (0 go on, pass, to spend (time) to go on, move on
deber de + infinitive should! must must
Looking at these verbs and their usage in the Spanish, we may be able to draw a
certain correlation between the verbs that are completely losing their original meaning
12
when modifying the second verb and their flexibility to be used in expressions in Spanish.
There are, of course, many more PV's in Spanish, but these cover the large varieties that
exist, and are a good representation (as far as I know.) I have taken the liberty of
breaking these verb phrases up into 3 tiers. In the fust tier, the dotted~lined box, the
periphrastic meaning is drastically different from the non-periphrastic one. In the second
tier, the lined-box, the periphrastic meaning carries much the same meaning, or is
figuratively the same as the non-periphrastic one. Finally, in the third (and smallest) tier,
the periphrastic meaning is practically the same as the non-periphrastic meaning. I have
broken them up into these tiers to show the semantic differences that can occur with these
special phrases. Instead of claiming that periphrastic phrases include a loss of meaning of
the first verb, arguing that there is a range from total loss of meaning to practically no
loss just might be fruitful. Later we will see that syntactically they behave the same,
whether they lose all meaning or not. Thus we may ask why a periphrastic verb must
lose its original meaning in order to be labeled periphrastic if it meets all of the syntactic
tests? And why should a different category of verbs be named when simply looking at
the semantic reasons for changes that do or do not occur can accow)t nicely for the losses
or lack thereof?
Certainly some sort of pattern can be noted. Those in the dotted lines are verbs
that are noticeably flexible in the language. That is to say, these verbs either have several
different meanings, or they can be used in slang expressions. They are very commOll
verbs used in all registers of language . When we look at the second tier, we find much
the same result. These verbs do seem to have several meanings, though their use in fixed
13
expressions may be slightly less common. And finally in the third tier, we find verbs that
do not have a variety of meanings, and are not flexible to the extent that they can be used
. . In expressIOns.
To put these observations to use, let us now look at some verbs that cannot occur
in periphrastic phrases. It is interesting to note, looking at the chart, the absence of high
infonnation verbs, and of verbs used in the most elegant of speech registers. What does
this mean? We can look, for instance, to the three verbs poner, meter, and coloear. All
three of these verbs can mean 10 put or fO place. PoneI' is used in everyday language, and
also carries a variety of other meanings. Its usage can be very colloquial, and it is not
considered at all eloquent. Meter also has a variety of meanings, but when used formally
as to put or to place, it is a bit more eloquent thanponer. Coloear is very specific,
meaning only 10 put or 10 place (something somewhere.) It is used in a highly eloquent
speech, and is not used as a periphrastic verb.
Another instance that is interesting is the difference between andar and caminar.
Both mean to walk. Caminar is not used in expressions, and it is not used in slang. It
cannot be found in a periphrastic phrase. As we have already seen, andar is used in
expressions, and in slang. And of course, it is used in Periphrastic Verb Phrases.
We can see now that periphrastic verbs tend to be those with multiple meanings in
non-periphrastic phrases, or verbs that can be used in expressions, or in less rigid social
environments. It is especially interesting to note the modal verb deber and its
periphrastic use. In most languages, the episternic and deontic use of this modal are the
same syntactic structures, but elegant Spanish draws a line between the two. The
14
epistemic use is a periphrastic phrase, whereas the deontic use is the modal alone. For
instance, in example 11 below, the periphrastic phrase gives a different meaning than the
modal alone would give, follow the most elegant Spanish register:
11. Sacas muy buenas notas,iy no estudias! Debes de ser muy inteligente.
You get good grades and you don't study. You must be very intelligent.
The verb deber as a modal, non-periphrastically, would be used as in an instance like
sentence 12:
12. Debes estudiar mucho; queremos que saques muy buenas notas.
You must study alo!; we want you to get good grades.
The difference between these two forms of deber is rather important, and it might
be an error to say that this verb realJy carries the same meaning periphrastically that it
carries originally. I have put it in tier 3 nonetheless because these two fonns have
become practically interchangeable for a phrase like II, the epistemic use, in Spain today.
While 13 is not considered correct, many people see no difference between saying 11 and
13:
13. Sacas rnuy buenas notas, ' y j no estudias! Debes ser muy inteligente.
You get good grades and you don't study. You must be very inteligent.
However, nobody would make the mistake of confusing the periphrastic phrase
with the modal in 12. The deontic use of the modaJ alone is not interchangeable with the
periphrastic phrase, so one would probably never encounter a sentence like 14:
14 . *Debes de estl.ldiar mucho; queremos que saques muy buenas notas.
Trying to say: You must sll.ldy a lot. We want you to get good grades.
15
The difference between the periphrastic phrase and the modal verb is not marked
in English, as we see from the translations above. The periphrastic phrase does carry a
different meaning though. To wrap up this section, we can note that while linguists do
not agree on what detennines a periphrastic verb phrase, all possible periphrastic phrases
do share one thing in common: they all are of verbs that have multiple meanings, are
colloquial, or are not high information verbs. This could be enough of a semantic tie to
ignore whether the verb completely loses its non-periphrastic meaning or not.
D. Translating Periphrastics
Translating periphrastic verb phrases can be especially difficult. The real
question, for at least some periphrastics, is whether you can translate them using the same
method from context to context, or whether each context calls for an individual treatment
of each periphrastic phrase. In other words, do they add the same basic meaning to the
verb they modify) no matter what the context? Or does the alteration of meaning also
depend on the context? To answer these questions, we will focus on one periphrastic
phrase that does not have an equal periphrastic counterpart in English. The periphrastic
verb wi1l be used in several different contexts, with several different verbs. OUf
periphrastic phrase will be llegar a + infinitive.
According to Jesus Fernandez, !legar a can indicate conseguir (to achieve), or a
nuance of "ir hasta el extrema" (to ga to extremes.) In addition, he suggests that it
sometimes adds a feeling of " por fin.''' (Finally!!) With no, it expresses a
disappointment, "sin realizacion" (without realization.) From Fernandez's explanation
16
of what !legar a can add to an infinitive, we might already suspect that this PV cannot be
translated the same way all of the time. In examples 15-17, we have 3 phrases containing
the periphrastic phrase lIegar a. The translation underneath each one is not complete.
The would-be translation of the PV llegar a has been Jeft blank, to look at possible
translations of the periphrastic phrase.
In these three examples 15-17, there are many periphrastic phrases. Focusing in
on !legar a, there are four occurrences. We will go through each occurrence now, and try
to figure out just what the periphrastic verb is adding to the main verb.
15 El canciller aleman, Helmet Kohl, afmno que "el combate fue extraordinariamente duro" y
que lJeg6a dudar de la posibilidad del acuerdo."
El Euro nace con Dolor Xavier Vidal-Folch El Pais
The German chancellor, Helmet Kohl, affirmed that "the combat was extraordinarily difficult"
and thal ______ the possibility of agreement."
In example 15, lleg6a dudar is our PV. Dudar means to doubt. We know that
this PV can carry a meaning of realization; here it seems to carry a measure of starting or
beginning (to doubt.) The actualization of doubting is what is important; the fact that he
doubted at all is stressed by Ilegoa. To capture this, we could maybe say
15.a He actually doubted the possibility of an agreement.
Because !legar a also indicates, in this instance, some sort of beginning of doubting, and
actually may not express this, a better translation may be:
IS.b The Gennan chancellor ... and that he actually began to doubt the
possibility of the agreement.
17
Actually should be maintained in this translation in order to keep the sense of
stressing that he shouldn't have had to doubt the possibility, or the fact that he had to at
all is important. The actualization of doubting is stressed.3 Of course) it's really difficult
to say what the best translation is, but 15.b seems to capture it pretty well. Let's look at
some other examples:
16 ... nos caJlabamos a escuchar como embobados. pero cuando mvimes mas arran que,
por el vino y Ja conversacion, nos liamos a cantar en rueda y, aunque nuestras voces
no eran demasiado templadas, como lIegaron a decirsc cosas divertidas, todo se nos
era perdonado ...
La Familia de Pascual Duarte Camilo Jose Cela
... we quieted down like idiots to listen, but when we got started up by the wine
and conversation we gathered in a circle and began singing without direction,
and even though our voices weren't really in tune, since they ____ _
funny things, we were forgiven all.
In example 16, our PV is llegaron a decirse. Decir means to say, and the
reflexive here indicates that the saying occurs among them. It acts as a sort of passive
construction and puts cosas divertidas (funny things) in the agent role. Thus, before
translating llegar a, we should pay attention to the structure itself, and change the English
structure to:
16 a .. . since funny things were said .. .
Now we will focus in on what llegar adds to this translation. The stress is on how
the things were funny. Again, as in Example 15, we could probably use actually:
3 Another interesting side 110te is the choice of the adverb extraordinariamente by the German (non-native)
Spanish Speaker. The word has 9 syllables, and is quite a standout in the phrase. After reading the article through, it is quite a standout in the writing style.
18
16.b ... since funny things were actually said ...
This translation may be a bit awkward and hard to interpret. It does not really fit
in well with the context of the phrase. To find a way to say this in English is very
difficult. One native English speaker who has lived in Spain for over 10 years and speaks
the language very well suggested that the phrase just be translated as in 16.c:
16.c ... and since funny things were said .. .
There are certainly reasons to reject a translation like this one, even if finding
another is extremely difficult. Translation 16.c is ofthe phrase se dijeron cos as
divertidas and not of llegaron a decirse cosas diverlidas. In the end I would refer to 16.b
as the best translation. More than anything, this shows just how difficult translating PV's
can be. Another example we can look at which contains the PV llegar a two times,
shows even more problems wjth translating this PV:
17 Mucha me diQ que pensar, en muchas veces, y aun ahara mismo si he de decir
la verdad, el motivo de que a mi madre Hegase a perderle el respeto, primero, y el carino
y las fonnas al andar de los anos, mucha me dio que pensar. porque queria hacer
un claro en la memoria que me dejase ver hacia que tiempo dejo de ser una madre
en mi corazon y hacia que tiempo lIeg6despues 8 convertirseme en un enemigo.
La Familia de Pascual Duarte
Camilo Jose Cela
[t gave me a lot to think about, times over. and even now if I must tell the truth,
the reason why I _ ____ respect and the caring ways for my mother over
the years, it gave me alot to think about. because I wanted to clear my memory so
that r could see when she quit being a mother in my heart and when she
_______ in an enemy.
In example 17 we have llegar a two times: llegase a perderle (perderle means to
lose plus the object, here being mi madre or my mother,) and lleg6 ... a convertirseme
19
(convertirse means fo convert with its reflexive, and the me refers to the speaker.) In the
first instance, Pascual Duarte is saying that he lost respect for his mother. Before adding
the periphrastic modification, a usual translation might be:
17.a .. .the reason why I first lost respect for my mother, and over the years the caring ways .. .
The modification of !legar a here is to stress losing. I'm not sure that actually
quite captures the meaning of the phrase, because it isn't stressed that its odd that he lost
respect at all, but rather that somehow he lost it. For this reason, even though actually
could probably be used here, I would have a preference for a translation Jike 17. b:
17.b ... the reason why [ ftrst somehow lost respect for my mother, and over the
years the caring ways ...
In translation 15.a, we included actually began in place of the periphrastic phrase.
Here again there is some hint of focus in when he lost respect, but this time ir comes from
the context, and the word first. The beginning in began to doubt also comes more from
the context than the periphrastic verb itself, probably. It is difficult to separate the two.
Perhaps it is not necessary.
In the second llegar PV, llego (despues) a convertirseme, we have the reflexive se
(referring to the one who converts), and the clitic me, referring to Pascual, the speaker.
Before the periphrastic modification, the translation I would propose is:
17.c .,. and when she turned me into an enemy ...
20
We could say converted me into an enemy, but this would probably not be an
elegant translation. Convert doesn't fit the tone of the text; it's too stuffy. As for the
addition of !legar a here, again it stresses the actualization of her making him an enemy,
and, again actually could be used in the translation.
17.d .. . and when she actually turned me into an enemy.
After looking rather quickly at these examples, one may want to say that llegar a
simply adds actually to the sense of the verb. However, by looking closer at the texts we
have seen that it is not that simple at all. It is really difficult to say whether or not each
translation with this verb really needs to be treated individually. When llegar a does not
follow no, or is not followed by ser, however, actually may be a decent translation.
There will always be exceptions though. Example B was rather difficult to translate.
Perhaps actually is not the best word choice. And actually certainly does not work with
ser. Llegar a ser is always translated as fo become, as seen here in example 18:
18 EI Ileg6a ser presidente
He became president
In addition, we can see with examples 19-21 below that no llegar a is different.
Actually does not translate well in all of these exampJes:
19 Queda ver la pelicula, perc no lIegu{ a verla.
I wanted to see the movie, but I ___ _
21
This phrase was given to me as an example of the frustration no llegar a can
show. I was told by a Spanish professor (Jesus Fernandez), as he stomped his foot in
disgust for not having had the chance to see the movie, that he would translate this phrase
as in 19a:
19.a I wanted to see the movie, but daggonit, I just didn't get to see it!
Because, in this case, daggonit is not really used that frequently, and because
more than anything daggonit is out of place, a translation like in 19. b, in my judgement,
would suffice:
19.b I wanted to see the movie, but I just didn't get to see it.
Comparing this translation of no Ilegar a with our earlier translation of llegar ~
we see a difference in the nuance that no alone couldn't add. However, with the
periphrastic phrase, no gives a sense of frustration that actually just does not give.
Compare translation 19.b with just to 19.a, the translation using actually instead:
19.c I wanted to see the movie, but I didn't actually get to see it.
Though I may be wrong, I believe that actually hints that the speaker expected to
get to see the movie (along with wanting to get to see it.) Llegar a does not, to my
knowledge, give any hint of this . In addition, actually does not show any great
disappointment on the part of the speaker. Another big difference between 19. b and our
earlier translations of the periphrastic phrase llegar a is the usage of to get to. What
22
would happen if we took it out of translation 19.b? Would the translation still be good?
We can see this result in 19.d:
19.d I wanted to see the movie, but I just didn ' ( see it.
19.d, like 19.c, does not express the disappointment or frustration at not seeing the
movie. Thus, the combinationjust + no! + get to (19.b), is probably the best translation
for 19.
Given another context with no lIegar a, would the same translation fit? Example
20 is from another passage in La Familia de Pascual Duarte:
20 Yo andaba preocupado y como pensativo, como temoroso del paso que iba a dar --
casarse es una cosa muy seria, ;que caramba! -- y momentos de Oaqueza y desfallecimiento tuve,
en los que Ie aseguro que no me falto nada para volverme atras y mandarlo todo a tamar vientos,
cosa que si no lIegu{ a hacer fue por pensar que como la campanada iba a ser muy gorda y, en
realidad, no me habia de quitar mas miedo ...
La Familia de Pascual Duarte, 82 .
Camilo Jose CeJa
I went around worried, and somehow pensive, apprehensive about the step T was about to
take -- to get married is a very serious thing, geez! -- and r had moments of weakness and
faltering, moments in which I assure you it wouldn't have taken much to make me tum back and
tell them all to get lost, which ifl , it was because I was thinking about how the bells
were going to ring out, and really, that took away enough of the fear.
Si nunca llegut'a hacer is a very interesting phrase to translate. Using our
previous combination, we would get a translation like 20.a:
23
20.a something that if I just never got to do,
This sounds like he really wanted to do it. From the context we can draw that that
is not necessarily the case. In fact, for the flow of the context, our earlier translation of
actually may be the best:
20.b something that ifl never actually did it...
This translation expresses that he thought a lot about doing it, but he never
actually did. The real sense of the passage is captured by this translation through the
word actually. Because for as much as he thought about it, he did not actually do it. And
that's what the phrase captures. 20b captures that well. Thus, no llegaf a does not
always indicate a frustration for not having gotten to do something. It can also indicate
the counterpart of llegar a, by simply meaning not actualized.
For almost any context, this periphrastic phrase has to be treated individually in
translation. There seems to be no lone method of translating it. This occurs with many of
these Spanish periphrastics, and is one of the main reasons so many linguists are studying
just what is beneath each phrase.
III. V IP Structu res
As we mentioned before in earlier sections, PV's include two basic types of
structures. In this section we will look at the V IP periphrastic structures and their
24
syntactic characteristics, including the types of movement they allow. The V IP
structures are those that have a periphrastic verb and either the past participle or the
-NDO form of another verb. Some possible V IP periphrastic phrases are given below in
sentences 21-23:
21 Este chico anda metilndose con la gente.
This boy goes around messing with people.
22 Yo andaba preocupado.
r went around worried.
23 a. Vaya ir yfndome
b. Me lIoy a ir yendo.
I'm going to get going.
21 should look familiar, as we saw it earlier. In 21 we have the V (anda) IP
(metiendose con la genIe) structure where the IP verb is the progressive (or -NDO) form.
In addition, we see the reflexive se adjoined to the end -of the progressive. Later we will
look at movements of this se, and of c1itics that can adjoin to the progressive.
22 gives us an example of a V IP structure where the second verb is the past
participle. The past participle in 22 agrees in number and gender with its theme yo. Tills
phrase may also look familiar, as it was taken from an earlier example which was chosen
from a passage in La Familia de Pascual Duarte.
23 is an especially interesting structure. First we have the V CP structure of voy a
ir, and in addition we have the V IP structure of ir yendo. The reflexive me can be either
attached to the progressive (as in B) or raised up in front of the first verb phrase ofthe
structure (as in A.) This first verb construction (voy a ir) is the CP structure we will later
25
explore. For now, lets look at the deep structure and surface structure of the V IP
phrases, with their lexical entries:
A. Lexical Entries and Sentence Structures
First we'll look at the lexical entries of our verbs. In sentence 21, we have andar
and melerse. In giving the entries for periphrastic verbs, we can either give lexical entries
for verbs when used periphrastically and non-periphrastically, or just when used
periphrastically. All that really interests us are the periphrastic uses when the verbs are
periphrastic. Thus, when periphrastic, we will only deal with them. This means we'll
ignore uses such as:
24. Este chico anda por la calle.
This boy walks the road.
Purposefully overlooking uses like these, andar takes only an IP. The verb of this
IP can be in either the progressive or the past participle fonn . No other fonn works.
Thus, the subcategorization frame for andar would be as given in 25:
25. Andar, V [_ IP [+ finJ]
The IP phrase it takes must be finite. As for melerse, in this instance, it takes a PP
sister where [con] is specified . So it's subcategorization frame is shown in 26:
26. meterse, V [_ PP [con]]
Melerse can also take a CP sister as a periphrastic phrase (meterse a), but the
periphrastic meterse is not the one used in sentence 21, nor subcategorized for in 26.
26
Given these subcategorization frames, we can see a sort of basic surface structure of 21 as
represented below in 27:
27 IP
~ NP l'
~ I[+finJ VP
. I V'
~ V IP
/--
I
V
VP \
V' ~
PP
27 is not a good representation of our example yet. We have to decide on a few
issues. Where does yo start out in the deep structure? As a contentful, overt NP, we
know it must get case. Where does its case come from, and where does its theta-role
/ come from? I now propose what Femandez de Castro (I 990) sought to prove about
periphrastic structures: the periphrastic verb does not assign the subject's theta role (nor
its case.)
If we accept Ferncfudez de Castro's claim, it is only natural for us to now propose
that the subject starts out in the Spec of the lower IP. There it gets its theta role from the
main verb. The subject then moves up to the spec of the higher IP clause in order to
fulfill the Extended Projection Principle. Therefore, the periphrastic verb itself does not
27
assign an external theta role. It does, however assign an internal theta role to its sister IP
(or CP) clause. This means the lexical entry of the periphrastic andar is as shown in 28:
28 . an dar, V [_ IP[+fin) ]
<6>
Andar assigns one theta role, and that is to its IP phrase. Sentences 21 and 22
should have the same basic structure, even though their IP phrases may differ. Because
sentence 22 is very interesting in that the past participle must concur in number and
gender with the subject, and because the deep structure of these two constructions as
periphrastic verb phrases will be similar, we will examine the likely deep structure and
surface structure of 22. To do this, we will need the lexical entry for preocupar: which
would be as given in 29:
29. preocupar, V [_ (PP(de] ) ]
-eft, (&»
Preocupar assigns an external theta role. Optionally it takes a prepositional
phrase as its sister, to which it can assign an internal theta role. With this information, we
can construct the Deep Structure (DS) of sentence 22, given in 30:
28
30 IP
/ .............. l'
/'--. I[+tinJ VP
\ V' /~
V IP (<8>. (rom "''''lsi)
/"""-NP (~, fro," l"e'OCIlpaJ.lII '
I~ VP I
V' I
V
In this structure j Th NP yo gets its theta role from preocupado. The higher Spec,
IP is empty right now. According to the Extended Projection Principle, Spec, IP must be
filled at SS. Thus, yo moves up to the fIrst Spec, IP position, and leaves a trace behind.
According to Burzlo's Generalization, if a predicate does not have an external argument,
than it does not assign case. Yo is a contentful noun phrase, so it must get case. Thus, we
may be forced to asswne that, like wh-question words, in periphrastic verb phrases this
contentful NP that moves up can inherit its case from the trace it leaves behind . This is
not a really odd concept, since it does occur. In addition, the past participle must agree in
number and gender with this contentful NP, or more likely, share spec/head agreement
with the trace the NP leaves behind. The NP yo then moves up to the higher Spec, IP
position. And we have a surface structure (SS) like that given in 31:
29
31 IP /"'-..
NP j L'
~ I [+finJ VP
\ V'
/". v IP ~
4 I'
/'---I( +tin] VP
\ V' \
V
This is an example of a raising structure. If the V IP clauses are indeed raising
structures as presented in trees 30 and 31: then we would expect these structures to also
permit cHtie climbing. In addition, we would expect that the special characteristics of
raising structures that can be seen in wh~movement and passive structures will also be
present and important in periphrastic phrases of this structure.
B. Clitic Climbing in VIP's
The following data in 32~34 show how the direct object clitic, and the reflexive
can be attached to either the end of the progressive verb, or can be moved to the front of
the first verb of the verb phrase. In all of these, both a and b are good sentences, which
mean the same thing:
30
32. a. Va cambilndolo.
b. La va cambiando.
S/he is slowly changing it.
33 . a. Anda metitndose con la gente.
b Se anda metiendo con la gente.
S/he goes around messing with people.
34 . a. Vay a ir ytr;dame
b. Me voy a ir yendo.
I am going to get going.
In the following section, we will look at the basic structures underneath this
movement. The lexical entries for ir, cambiar, and meter are given below in order to
generate trees for these phrases. This time we have added the possibility of the CP to the
periphrastic ir. However, in the data above we also have the non-periphrastic ir in
sentence 34. Therefore, we have given the lexical entry for the periphrastic ir in 35, and
the non-periphrastic ir in 36:
35. ir, V [_ {IP[+tin)} )
{CP[-finJ}
<8>
36. ir, V [_ ]
<ft>
As can be seen with these lexical entries, the structure of the periphrastic ir is
really different. It can also take a CP, which we will later see. To be periphrastic,
however, it MUST choose between the IP and the CPo One of them is necessary. It' s
31
, -
internal theta role then goes to either its IP or its CPo The other lexical entries necessary
(before we see the structures) of these sentences are given in 37 and 38 below:
37. cambiar, V L (NP)]
~, (~»
3 8. meterse, V L ({Pp[conJ })]
<§,(.fl2»
Cambiar (which is non~periphrastic, of course) can optionally take an NP, to
which it then assigns an internal theta role. Meterse, non~periphrastic here, can take a PP
sister with the preposition con, to which it then assigns its internal theta role. We can see
through these examples that the V IP structures do indeed allow clitic raising. In trees 39
and 40 below we will see first the DS of32 (a and b, which share the same DS) and then
the SS of 32 a. After this, we will see the further movement that occurs between the DS
of 32 b (the tree given in 39) and its SS in tree 41. From the DS in 32, we see how the
clitic starts out attached to the progressive, and the subject, in this case pro, starts out in
the lower Spec, IP position. (Spanish is a pro drop language.) The c1itic itself is
co~indexed with some direct object, which is normally represented with PRO:
32
39. IP
/~ I'
~ I [+fin] VP
\ V'
~ V IP <&> (l'Iom "")
va NP~I' / ~
VP
\ I [+fin]
V'
v ~NP A I
pensando 10j PRO) < (» (rrom p ... lndo)
Between 39 and 40 we show the movement ofpro to the higher Spec, IP. This is
the SS of 32a, represented here in 40:
40. IP /~
pro; l'
/'--VP
\ V'
v~--------- IP _
I ~ va NP
\ l'
/-----I(+finJ VP
1 v' ~
33
v ~
pen sando 10' J
NP
\ PRO
j_
We know that in sentence 32b the cIitic 10 has moved up to pre-verb posistion.
Example 41 shows this further movement:
41 IP
N~~I' I ~
proj I[+fmJ VP
\
~~ V IP_
/" ~ 10j va NP l'
\ VP \
I [+finJ
V'
~ V NP
/'--.... \ pensando tj PROj <01:>
From examples 33a-b and 34 a-b, we can see that reflexives can also be in either
position. Again, this is a common trait of raising constructions. If this is common only
of raising structures, than there must also exist those that do not allow this movement.
Control stuctures don't allow this. We will later see instances where the clitic and
reflexive cannot be raised from their attached position at the end of the last verb of the
verb phrase to the position before the first verb of the verb phrase.
34
C. WH-Movement in V JP's
The following data show wh-movement in the V IP structure. Just as
clitic-raising is fine, so is the A' movement ofwh-questions:
42. a. Va hacifndolo.
S/he is slowly doing it.
b.C-Que va haciendo?
What is slhe (slowly) doing?
This certainly is not surprising, because we would expect a raising structure to
allow wh-movement out of the object position. In 44 we have given the lexical entry for
decir (not periphrastic):
44. decir, V [ _ {NP } CPP)]
{CP [que)}
< fu, &2, (93» Decir assigns an external theta role to its agent. In addition, it assigns a theta role
to its NP o.r CP. This verb can also optionally take a PP sister along with its NP or CP, to
which it can assign a third theta role. Thus, we have a surface structure of 43 a like in 45:
45. IP
~~ NP I'
I ~ proj I[+finJ VP
\ V'
~ V IP
~ anda NP l'
\ /"---..... I[+fin] VP
\ V'
35
v NP
~ diciendo
The DS of the sentence 43 b would then be as given in 46:
46. CP [+qu]
A C' ["'qu]
~ IP C (+qu]
~ I'
~ I [+fin] VP
~ V'
~ V IP
~ anda NP I'
~ proj I[+fin] VP
\ V' ~
V NP
diciendo /" que
10 mismo
Finally, in 47 with the A'-movement of the NP qut, and the movement of pro
from the lower Spec, IP to the higher Spec, IP, we generate the SS of 43b as given in 47:
47. CP [+qu]
~ NP C' [+qu]
36
~ quej C [+qu) IP
~ NP l'
\ ~ VP \ V'
~ V IP ~
anda NP I'
\ ~ t-, VP
t V'
/'-... V NP
diciendo t J
The wh-movement that occurs in 47 (sentence 43b) is fine, just as we would
predict of a raising structure. The trace left behind gets both its theta role and its case
from the verb decir, which is also fine. If we are to believe Fema;;dez de Castro ' s
assertion that the first verb of the periphrastic phrase gives neither case nor a theta role,
the progressive and the past participle must at some point be capable of giving the trace
of the NP that has moved up its case. Then the NP must inherit its case from that trace.
D. Passive Movement in V JP's
One of the key characteristics of raising strucutres involves passive movement. In
this section we will explore the passive movement of a periphrastic V IP structure and
37
compare that with the passive movement of a control V IP structure. TIlls will then help
confinn OUI lexical entries and earlier structures of periphrastic verbs by showing what
happens when a verb is not of this type. Consider the following example sentences
46-49, which are examples presented by Fernru;.dez de Castro (1990)4:
46. EI gobiemo va Iiberando a los presos.
The government is slowly liberating the press.
47. Los presos van siendo liberados por el gobiemo.
The press is slowly being liberated by the government.
48. EI gobiemo acerttiiberando a los presos.
The government succeeded in liberating the press.
49. Los presos acertaron siendo liberados par el gobierno.
The press succeeded in being liberated by the government.
46 and 47 are logical equivalents, but 48 and 49 are not. We will now explore the
reason for this. The underlying structures of 46 and 47 are the same, whereas the
under1ying structures of 48 and 49 are different. 46 is an example of a periphrastic
construction, and as we have indicated before, a raising structure. 48 is an example of a
control verb (acertar), and for this reason, a control structure. The first step to showing
this underlying difference is to show the difference between our verbs used . The lexical
entries of ir, liberar, and acerfar are given in 50~52:
50. ir, V [_ {VP [+modal] }]
{CP (al }
{IP [+fin] }
<e>
4 The translations of these sentences are 5tiJ I my own.
38
51. liberar, V [_ NP] ~,e2>
52. acertar, V [_ (VP [+modal))]
(IP (+finJ)
<ft .. 82>
Ir, as a periphrastic verb, may choose among a VP, CP or IP sister. It assigns an
internal theta role to its CP or IP sister. Liberar assign both an external theta role, and an
internal theta role (to its NP.) Acertar assigns both an external and an internal theta role.
The internal theta role goes to its IP. Just as in our earlier trees of periphrastic phrases,
the NP moves up from the lower Spec, IP position, and leaves behind a trace. The SS of
46 would be as shown in 53:
53. IP
~~ NP I'
~ el gobiernoj I [+fin] VP
\ V' ~
V IP (<II>. (ro", '''')
~--va NP I'
\ ~ tj (<"&>. (,om 11b.:"-"1<l<,) I[+fin] VP
) V' ~
V NP
liberando a los presos«1I'> (rom Illx:raMv)
39
The reason 47 is the logical equivalent to 46 is because, in its DS, the same theta
roles get assigned to the same NPs by the same verbs. The only fundamental difference
is the movement between DS and SS. The DS of 47 is shown in 54:
54.
V
van
VP (
V'
~ VP \
V'
/~ V IP ~
siendo I'
~ I[+fin] VP
~ V' P P (<:iI>. r(Om lil><ro')
~ ~ V NP «!P. (ro," Jib, ,,,,,) por el gobierno
b::.-liberados Ios presos
The previously external theta role of liberar is now assigned to el gobierno
through the prepositional pJu-ase. The internal theta role of liberar still gets assigned to
los presos. Before SS, los pres os moves up to the lower spec, IP position. There we find
Spec, head agreement, and then it moves again, leaving berund a trace. It must stop off
there in order to fulfill the Extended Projection Principle. That is to say, all Spec, IP's
must be filled. Because it stopped off there, liberados must concur in gender and number
40
with los presos. At this point, there is still an empty Spec, IP position. The next
movement of los presos is to that higher Spec, IP position, where it remains at SS. This
movement is just like in our earlier periphrastic structures. Thus, the SS is as shown in
55 :
55. IP
/~ NP l'
b. ~ los presosi I [+finJ VP
t V' ~
V VP
\ van V'
~ V IP ~
siendo NP I'
/~ I[+finJ VP
~ V' PP «6>. from Jjb.,."""')
~ ~>-V l
liberados
NP ("f1>'(ro)nfj/)<Nx/"'J por el gobiemo \
~
Thus, in this raising structure 46 and 47 are logical equivalents because the same
theta roles are assigned to the same contentful NP's. To compare, let us now look at the
DS of 48 and 49 in 56 and 57, respectively:
56.
NP I'
41
el gobiemo, I [+finJ VP I
V'
V~IP /-----
acerto NP l'
/~ PRO; I[+finJ VP
\ V'
V~NP ~
liberando a5 los presos
The major difference between 56 and 53 is the PRO in 56, coindexed with el
gobierno, and the trace left behind by el gobierno in 53. In 56, el gobierno starts out in
the higher Spec, IP position. This is because acertar assigns an external theta role.
Liberar also assigns an external theta role, so it is necessary to have a contentful NP in
the lower Spec, IP position that can absorb that theta role. PRO counts as a contentful
NP, and absorbs that theta role. It is co-indexed with el gobierno (as mentioned before.)
This major difference in the structures accounts for the reason that 48 and 49 are not
logical equivalents. 56 (48) is an example of a control structure. Next, in 57, we have the
DS of 49:
57. IP
NP l'
los preSOSj I[+fin] VP
5 This a is the personal a in Spanish. It is a preposition, but since this does nol really make a difference in our analysis, for now we will treat it as part of the NP.
42
V' ~
v
acertaron
VP t
V' ,..,....-~
V IP
/"----siendo NP VP
/"'----PRO; V'
\ V
liberados
pp
por el gobiemo
El gobierno, in 56, gets its external theta role from acertar . In this tree it gets its
theta role from liberar. Los presos gets the external theta role from acertar. This
accounts for the difference in the meanings of 48 and 49.
Thus, we can see that raising predicates have a different behaviour in passive
. movement than control structures. And periphrastic phrases with the V IP structure are
indeed raising predicates. We have yet to examine the V CP construction, which is for
the next section.
IV. V CP Structures
We have seen a couple of instances of V CP constructions to this point, but in this
section we will concentrate solely on their structure. in the sentences 58-60 given below,
we have several different V CP constructions:
58. Voy a ponerrne a rrabajar.
I'm going to set myself down to work.
43
59. Acabo de hincanne de comer chocolate.
I have just eaten chocolate til I popped.
60. Iba a dar un paso.
I was going to take a walk.
61. Acabo por comerlo.
I ended up eating it .
58 and 59 are interesting because they combine periphrastic phrases. In them we
find a verb plus CP which contains a verb plus a CPo 60-1 show periphrastic V CP
constructions in which there is only one V CP construction. In addition) in 59 and 61 we
see the verb phrases acabar de and acabar por. This shows us that different
complementizers do change the meaning, and that the verb subcategorizes for its
complement. A sentence like 62 below is bad because aca~ar does not subcategorize for
the complementizer a:
62 "Acabo a comerlo.
I finish to eat it.
Just as we saw c1itic rasising, wh-movement, and passive structures in the V IP
structure, we will explore these movements in the V CP structure.
A. Lexical Entries and Sentence Structures
In V IP's) Fernandez de Castro (1990) asserts that the theta role of the NP in the
highest Spec, IP is assigned by the non-periphrastic verb. He claims the same holds true
for the V CP structure. As one might already have noticed, the verb of the CP phrase is
44
never finite) so the periphrastic verb must subcategorize for a non-finite verb. This means
we must assume that CP can inherit [-fin] from its IP clause below.
Looking back now at example 58, we have two periphrastic verbs that
subcategorize for a CP [-fin], and our lone verb, trabajar. The lexical entries necessary
to construct a tree are given in 63-65:
63. ir, V [_ {IP [+fin]} ] {CP [a, - fin]} <9>
64. ponerse, V [_ {CP [a, -fin]} ] <8>
65 . trabajar, V [_]
~
Periphrastic ir, as seen before, can take a finite IP, or it can take a non~finite CP
with the complement a. It assigns an internal theta role to its IP or CP clause. Ponerse,
as a periphrastic, takes a non-finite CP with a as its complement. Finally, trabajar
assigns an external theta role and does not have an internal argument. Given these lexical
entries, we can generate the OS of 58 as in 66 :
66. IP
~ I'
~ 1[+£0] VP
\ V'
~ V CP «8>. ('0"' voy)
\ C'
/~ C IP
45
a l'
~ I [-fin] VP
\ V'
~ V CP (<4>.r",mpoJ/m ... )
C' /~
C IP \ ~ a NP l'
pro(<9:>.f(Oln lra""jUT) I [-fin 1 VP \ V'
I V
trabajar
In this DS, the fIrst CP gets ir's internal theta role. The second CP gets
ponerme's internal theta role, and pro gets the external theta role of trabajar. For the SS,
pro eventually moves up to the highest Spec, IP position. Along the way, however, it
must stop off at the Spec, IP of the higher CP phrase. This fulfills the Extended
Projection Principle, and each trace is antecedent governed. Finally we have a SS as
shown in 67:
46
67. IP /~
~ I [+finJ VP
\ V' ~
V CP «e;>. from >'Of)
I c' ~
C IP /-----a l'
A I [-fin) VP
\ V'
/'--.
c' r---C IP
~ a NP I'
\ A ~(<11>. r""" ,,"!.oM I [-fin] VP
"" V' \
V
trabajar
We assumed that the NP subject got its case from the progressive or past
participle in the V IP structure, because according to Burzio' s Generalization, if a verb
does not have an external argument, than it does not assign case. Thus, we must again
47
assume this pro somehow inherits case from the foot of its chain (of traces.) This also
means that a non-finite verb assigns its case. Again, we have a raising structure.
In the next few sections, we will examine the types of movement that occur in the
v CP periphrastic verb phrase. Oddly enough, the addition of the CP will not change the
types of movement that can occur.
B. Clitic Climbing in V CP's
The following data in 68 - 71 represent some phrases with clitics and reflexives
arranged in their possible positions for the sentences. A and B mean the same thing, and
both are good sentences:
68. a. Voy a ponerme a trabajar.
b. Me voy a poner a lrabajar.
I am going to set myself to work.
69 . a. Voy a ir yfn'dome.
b. Me voy a iT yendo.
J am going to get going.
70. a. Vas a tener que hacerlo.
b. Lo vas a tener que hacer.
You are going to have to do it.
716. a. Vay a hacerlo a Paris.
b. Lo vay a hacer a Paris.
I'm going to do it in Paris.
6 Example 71 a and b were taken from Fernandez de Castro, 1990. The translations are my own.
48
In examples 68 and 69, we see that the reflexive can move up over the CP phrase.
In 70 we see that the clitic 10 can climb over even two periphrastic CP's. Lets look at
how the raising of the clitic occurs. To do this, we need the lexical entry for (ener
(periphrastic) and hacer. We already have the lexical entry for ir (63), which is repeated
here as 72 for convenience:
72. ir, V [_ {IP [+fin]} ]
{CP [a, - tin] }
<9>
73. tener, V [_ CP [que, -fin] ]
<9>
74. hacer, V [_ NP] <!t1,62>
Tener, like all periphrastic verbs, assigns one internal theta role to its CPo Unlike
other periphrastic verbs, it subcategorizes for the complement que (that), but like all
others that subcategorize for a CP, it requires a ~ft.nite CP predicate. Hacer takes an NP
sister and assigns its internal argument a theta role. In addition, it's not periphrastic, and
it assigns an external theta role. The DS of sentences 70 a and b is given in 75:
75. IP
/"'-. l'
I [-fin] VP \
~ V CP I I
vas C'
49
~ C IP
\ /\ a I'
I'--. I [-fin] VP
\ V' /~
V CP I
tener C' ;-------. C ~
que NP
pro I [-fin] VP
\ V'
V~;; ~
hacer IOj PRO· J
The little pro moves up first to Spec, IP of the lower IP clause, then to the higher
Spec, IP. It gets its theta role from hacer, and somehow inherits case from its trace in the
SS. The clitic to can climb up over all of the CP's. (It is difficult right now to see just
how that affects or does not affect OB Theory. This is definitely something for further
research.) The surface structure of 70b would be as given in tree 76:
50
76. IP /, NP l'
J /'
pro; I [-fin) VP
\ V' A
V CP
1\ \ c'
/\ C IP
~ NP~I' \ ~
pro; I [-fin] VP
/ V'
I~ V CP
\ l tener c'
/'''- .
C " 'IP \ /'-.-
que NP I'
~ I [-fin] VP
\ V'
~ V NP ~ \
hacer tJ PRO· J
While we can definitely see that this clitic raising happens, and we know this is
characteristic of raising constructions, we do lack an explanation of this. It is very
important, as it may pose a problem for our assumptions. As mentioned before, this is
51
something to be further researched and thought out. For now, we will accept this as
further proof that V CP's too are raising constructions. Further proof of this is the
example in 77. also taken from Femaridez de Castro (1990). In this sentence, clitic
climbing is not possible:
./ 77. a. Voy a Pans a hacerlo . ./
b. *Lo voy a Paris a hacer.
I am going to Paris 10 do it.
It is important first to see that the ir in 77 is not periphrastic. Thus, it does assign
an external theta role, and case (to pro.) The a in a Panj is a preposition. A in a hacerlo
is a complement. The ditlc cannot climb over these two phrases. This is not a raising
construction. There is more to explore about why the clitic can climb in the raising
structures ofPV's and not structures like 77. Again for now we will just take this as
confirmation that the periphrastic phrase is a raising construction.
c. Wh-Movement in V CP's
TI)e following data reflect wh-movement in the V CP structures. As can be
expected, in an instance where clitic climbing is permitted, so is wh-movement. When
wh-movement is not allowed, neither is the raising of the clitics. One can compare 78-80
below with earlier examples to see this is true:
78. a. Vas a tener que hacer algo.
You are going to have to do something. ,-
b. ~ Que vas a tener que hacer?
What are you going to have to do?
52
/ 79. a. Voya hacerlo a Pans.
I'm going to do it in Paris. ,/
b. c..Qut yoy a hacer a ParIs?
What am I going to do in Paris?
" 80. a. Voya Pans a hacerlo .
I'm going to Paris to do it.
b. * ~ue voy a Parts a hacer?
Trying to say What am I going to Paris to do?
Immediately, 80 jumps out as a non-periphrastic verb structure, and we also note
the A-bar movement of wh-phrases is not allowed. In 78 we have A-bar movement over
two CP clauses. For the interest of seeing how the wh-movement occurs, let's look at the
DS and S S of 78 b. We have our lexical entries for these verbs in 72-73 of the last
section. The DS of 78b is shown in 81:
81. CP [+qu]
A C'[+qu]
G C[+quj IP
1\ r I'" I [.fin] VP
\
/~ V CP
\ \ vas C'
~ C IP
53
/\ a I'
I~ I (-fin] VP
/ V'
A v CP
\ \ tener C'
C~lP \ ~
que NP I'
/ ~ VP
\ pro I [-fin]
V'
v~P hacer
-,.../ que
In this tree, pro gets its theta role from hacer, which also given an internal theta
role to que. Vas gives a theta role to its CP, and tener gives a theta role to its CPo For the
SS, pro moves into the lower Spec~ IP, leaves a trace, and then moves to the higher Spec,
IP position. Qu{ also moves, obviously. It cannot move directly to the highest Spec, CP
position. The trace would not be properly licensed, because it would not be in the closest
possible A' position. Thus, gut first moves up to Spec, CP that is sister to the verb tener.
It leaves a trace there and continues to move up. It cannot yet go to the Spec, CP [+qu]
position, because this second trace of the chain would not be properly licensed. There
exists another closer A' position: Spec, CP (the CP that is sister to the verb vas.) After
54
moving to this position, que then moves fmally to the highest Spec, CP: iCs goal to form
the question. A tree reflecting this movement, the SS of 78b, is given in 82:
82. CP [+qu]
A NP
\ /
Quei
C' [+qll]
~ C[+qu] IP
A NP l'
\ /".... prOj I [-fin] VP
\ V' ~
V CP
\ /'.-. vas NP C'
I /"-.... tj C
\ a
IP
I~ NP r \ /"-.... tj I [-fin] VP
I
V I
tener
V'
A CP
/'--.... NP C' l /~ t; C IP , /~
que NP
/ tj I [-fin] VP
I V'
55
/ ~ v NP I J
hacer
Qulinherits its case from the foot of its chain. As given, wh-movement of
periphrastic V CP's is further proof that they are indeed raising constructions.
D. Passive Movement in V CP's
In this section we will explore the passive constructions of the V CP structure
opposed to a non·raising (or, in other words, a control) predicate. In the examples 83~86,
given next, we will consider passive movement, and then raising of the passive structure
in periphrastic phrases. Finally, we will see passive movement where raising brings
about different results in control predicates:
83. Este hombre va a causar problemas.
This man is going to cause problems.
84. Los problemas van a ser causados por este hombre.
Problemas are going to be caused by this man.
85', EI gobiemo desea afrontar cse asunto.
The government wishes to address that issue.
86. Ese asunto desea afrontado por el gobierno.
This issue wishes to be addressed by the government.
As with the V IP periphrastic structures, we immediately note that 83 and its
passive 84 are logically equivalent. Again we see a control structure (85) and its passive
(86) are not equivalent. The same analysis that we did in the V IP section about passive
holds true for the V CP structure. Because los problemas gets its theta role as the internal
1 Examples 85·86 are taken from Fern~dez de Castro (1990)
56
argument of causar in both 83 and 84, the two examples are logically equivalent. To
make a passive equal to 85, ese as unto would have to get an internal theta role from
afrontar. In 86 it gets an external theta role from desea. [n 87,90, we have our lexical
entries necessary for tree structure to explain this further. Periphrastic ir has been
repeated from 35 earlier for convenience:
87. ir, V [_ {IP (+fin]} ]
{CP [a,. fin] }
<&>
88. causar [_ NP] ~\, e2>
89. desear, [_ {IP [-fin] } ] {NP}
<fu,82>
90. afrontar, (_ NP] <it,92>
In 83, problemas gets its internal theta role from causar, and este hombre starts
out in the lower Spec, IP position where it gets the external theta role from causar. As in
all PV constructions, it then moves up ot the first Spec, IP position. 84 is the logical
equivalent to this because problemas starts out in the DS as the internal argument of
causar. It then moves up to Spec, IP of causar and we have spec/head agreement with
the past participle causado. Este hombre stilI gets the same theta role in 84 that it got as
the external argument of causar in 83 . The surface structure of 84, along with all of this
movement and theta role assignments, is reflected in 91:
57
91. IP
~~ NP r ~ ~ Los I [+fin] VP
\ problemas; V'
~ v CP (~from '<l> , \ van C'
/~ C IP
\ A a ~ I'
\ /"" ~ I [-finJ VP
\ V' /~
V IP
I /~ ser NP I'
I ~ t, I [+fioJ VP
/'--------V' PP <l6>
V/' NP~ 40~ \ I
causados ti este hombre
Passive movement occurs, then the raising that nonnaJly occurs in periphrastic
phrases takes los problemas up to the first Spec, IP position.
In order to get an equivalent passive sentence of 85, we would have passive
movement, but raising does not occur, because it is a control structure. 92 shows only
passive movement in 85, which would be a logical equivalent to it:
58
92. IP
NP~I' ~ /~
el I [+fin] VP \
gobierno V'
/~ V IP , ~
desea NP I'
l ~ ese asunto I (+ fin] VP
I
afrontados t I
In the actual passive structure of this sentence, we would find a CP (which may
exist here with a null complement) as the sister to the verb desear. In 92 el gobierno still
gets the external theta role of desea, and ese asunto gets the internal argument
structure/theta role from afrontar. In the passive movement, it moves to Spec, IP, sister
to ser, where spec/head agreement between the past participle afrontar and ese asunto
takes place. Finally, to finish its passive movement it goes to the next available Spec, IP.
In the DS of 85, we would find PRO in that spot. This is an example of passive
59
movement in a control structure. As we can see, passive movement in raising structures
is distinctive.
v. Conclusion
Periphrastic verb phrases are truly an interesting phenomenon in the language.
Whether the first verb loses its orginal meaning or not, the same syntactic characteristics
separate them from other verb structures. Both V IP and V CP structures are raising
predicates. Many linguists study them today because they are so very difficult to
translate. Because they are raising structures with CP's, it is also very interesting to see
just how Government and Binding Theory accounts for them. At this point, there are
many aspects left to be accounted for, from case assignment to localization effects. Using
our current assumptions of Government and Binding, we do at least know they are raising
constructions that allow clitic climbing, wh-movement, and have very distinctive passive
characteristics. The periphrastic verbs do not assign theta roles and case. This actually
only begins to explain their unique characteristics.
60
Bibliography
Cela, Camilo Jose . La Familia de Pascual Duarte. Literduplex, S.L. ed. Tregesima:
Barcelona, Espana. Noviembre, 1996.
Chomsky, Noam. Barriers. MlT Press. c. 1986.
Fente G6mez, Rafael; Jesus Fem~adez Alvarez, and Lope G. Feijoo. Peri}rasis
Verb ales. Madrid, 1987.
Fernandez de Castro, Felix. Las Per{jrasis Verbales en Espanol: Comportamiento
Sintactico e Hisforia de su CaracterizaciOn. Oviedo. Departamento de Filologia
Espanola. Espana. 1990.
Fernandez, Jesus; Rafael Fente and Jose Silas. Curso Intensivo de Espanol. Spain, 1990.
Gili y Gaya, Samuel. Curso Superior de Sintaxis Espanola. Ediciones Minerva, S. de
R.L. Mexico, D.F. 1943.
Gonzalez, Garcia Javier. PedJrasis Verbales. Prim era Edici6n: Sociedad General
Espcfuola de Libreria, S.A .. Espana, 1992.
Gomez Fuentes, Angel. AszCambiaraEspana: La Batalla del Mercado CornCn.
Grimshaw, Jane and Amlin Mester. "Light Verbs and a-Marking." Linguistic Inquiry.
Vol. 19, No.2. Spring, 1998 . 205-232. c. Massachusetts Institute of
Technology, 1988.
Havo,1. La ConstituciOn Temporal del Sintagma Verbal en el Espa1iol Moderno. 1998.
61
Moore, John. "Reduced Constructions in Spanish." University of California Santa Cruz.
c. June 1991.
Napoli, Donna Jo. "Semantic Interpretation Vs. Lexical Governance: Clitic Climbing in
Italian." Language. Vol. 57, No.5. 1981. P 841-887. University of Michigan.
Napoli, Dorma Jo. Syntax Theory and Problems. Oxford University Press. New York,
New York. c. 1993.
01bertz, Hella. Verbal Periphrases in a Functional Grammar o/Spanish. Editors: A.
Machtelt Bolkestein, Casper de Groot, and J. Lachlan Mackenzie. New York.
Mouton de Gruyter. 1998.
Rizzi, Luigi. Relativized Minimality. U.S.A., MIT Press. c. 1990.
62